Update app.py

app.py

@@ -9,17 +9,17 @@ from transformers import AutoModelForCausalLM, AutoTokenizer, TextIteratorStream
 
 MAX_MAX_NEW_TOKENS = 2048
 DEFAULT_MAX_NEW_TOKENS = 1024
-MAX_INPUT_TOKEN_LENGTH = int(os.getenv("MAX_INPUT_TOKEN_LENGTH", "4096"))
+MAX_INPUT_TOKEN_LENGTH = int(os.getenv("MAX_INPUT_TOKEN_LENGTH", "2048"))
 
 DESCRIPTION = """\
-# palmer-003
+# palmer-004
 """
 
 if not torch.cuda.is_available():
     DESCRIPTION += "\n<p>Running on CPU 🥶 This demo does not work on CPU.</p>"
 
 if torch.cuda.is_available():
-    model_id = "appvoid/palmer-003"
+    model_id = "appvoid/palmer-004"
     model = AutoModelForCausalLM.from_pretrained(model_id, device_map="auto", load_in_4bit=False)
     tokenizer = AutoTokenizer.from_pretrained(model_id)
     tokenizer.use_default_system_prompt = True